AI

Kim Kardashian says ChatGPT is her ‘frenemy’

Celebrities: Do They Look Like Us? At a vanity fair interviewKim Kardashian – who has been studying to become a lawyer – discussed her ‘toxic’ friendship with ChatGPT and admitted she failed law exams after it told her false information.

“I use [ChatGPT] for legal advice, so if I want to know the answer to a question, I take a picture, take a picture of it and post it there,” she said. ‘They’re always wrong. It caused me to fail the tests… And then I get angry and yell at it and say, “You made me fail!”

ChatGPT is prone to hallucinations, meaning that the LLM will sometimes create false answers, rather than admit that it cannot respond to a prompt with confidence. This technology is not necessarily programmed to know what information is ‘correct’ or not. Rather, it is trained on an unfathomably enormous amount of data and is pushed to predict the most likely response to an input, which may not actually be accurate. Some lawyers have been penalized for using ChatGPT when writing legal briefs, which is evident to those reviewing the briefs as they mention cases that do not exist.

Kardashian said she will try to appeal to ChatGPT’s emotions after it lets her down – but this is a futile plan, as ChatGPT has no feelings.

“I’ll talk to it and say, ‘Hey, you’re going to make me fail, how does that make you feel like you really need to know these answers?’” she said. “And then it says to me, ‘This just teaches you to trust your own instincts.'”

See also  The State of AI in 2025: Key Takeaways from Stanford's Latest AI Index Report

But while ChatGPT may not have feelings, that doesn’t mean we – humans – don’t have feelings about It.

“I’m constantly taking screenshots and sending them to my group chat, like, ‘Can you believe this b— is talking to me like that?'”

Source link

Back to top button